2022 56th Annual Conference on Information Sciences and Systems (CISS)最新文献

{"title":"Neural Bandits for Protein Sequence Optimization","authors":"Chenyu Wang, Joseph Kim, Le Cong, Mengdi Wang","doi":"10.1109/CISS53076.2022.9751154","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751154","url":null,"abstract":"Protein design involves searching over a large combinatorial sequence space. Evaluating the fitness of new protein sequences often requires wet-lab experiments that are costly and time consuming. In this paper we propose a neural bandits algorithm that utilizes a modified upper-confidence bound algorithm for accelerating the search for optimal designs. The algorithm makes adaptive queries as guided by the kernelized neural bandits. The algorithm is tested on two public protein fitness datasets, the GB1 and WW domain. For both datasets, our algorithm consistently identifies top-fitness protein sequences. Notably, this approach finds a diverse and rich class of high fitness proteins using substantially fewer design queries compared to a range of alternative methods.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130932373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multivariate Spatio-temporal Cellular Traffic Prediction with Handover Based Clustering","authors":"Evren Tuna, A. Soysal","doi":"10.1109/CISS53076.2022.9751165","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751165","url":null,"abstract":"We consider an RNN-based traffic volume prediction, which is a critical problem for network slice management and resource allocation in slicing-enabled next generation cellular networks. We propose to use a novel cost function that takes SLA violations into account. Our approach is multivariate and spatio-temporal in three aspects. First, we consider the effects of several other RAN features in a cell besides the traffic volume. Second, we introduce feature vectors based on peak hours of the day and days of the week. Third, we introduce feature vectors based on incoming handover statistics from the neighboring cells. Our results show about 60% improvement over MAE-based univariate LSTM models and about 20% improvement over SLA-based univariate models.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"54 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125968881","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Classification Predictive Modeling of Dyslexia","authors":"Yih-Choung Yu, Khaknazar Shyntassov, Amanuel Zewge, L. Gabel","doi":"10.1109/CISS53076.2022.9751182","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751182","url":null,"abstract":"Dyslexia is a reading disability that affects children across language orthographies, despite adequate intelligence and educational opportunity. If learning disabilities remain untreated, a child may experience long-term social and emotional problems, which may influence future success in all aspects of their lives. Early detection and intervention will help to close the gap between typically developing and reading impaired children in acquiring reading skills. We have demonstrated that animal models of dyslexia, genetic models based on candidate dyslexia susceptibility genes, and children with specific reading impairment show a common deficit on a virtual Hebb-Williams maze task. Since virtual maze task does not require oral reporting (rapid access to phonological processing) or rely on text, performance is not influenced by a potential difference in reading experience between groups. Although the correlation between dyslexia and the performance in the virtual Hebb-Williams maze task has been demonstrated, classification of atypical participants (i.e., dyslexic participants) through real-time observation of their performance on the virtual Hebb-Williams maze task is not feasible at this time. A computational model based on machine learning algorithms, that can predict reading ability based on maze learning performance, would enable real-time feedback of the performance in the form of at-risk percentages for reading. This paper presents the preliminary results of employing machine-learning based computational models to classify virtual maze performance on this task. Reading data and maze learning outcomes were analyzed from 227 school-aged children (8–14 years of age). Applying multiple variables, such as age and biological sex, into machine learning algorithms resulted in the prediction accuracy above 70%. Successful development of this predictive model would allow for early detection of risk for reading impairment, which can lead to early interventions to close the gap between typically developing and reading impaired children in acquiring reading skills.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127191542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decentralized Graph Federated Multitask Learning for Streaming Data","authors":"Vinay Chakravarthi Gogineni, Stefan Werner, Yih-Fang Huang, A. Kuh","doi":"10.1109/CISS53076.2022.9751160","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751160","url":null,"abstract":"In federated learning (FL), multiple clients connected to a single server train a global model based on locally stored data without revealing their data to the server or other clients. Nonetheless, the current FL architecture is highly vulnerable to communication failures and computational bottlenecks at the server. In response, a recent work proposed a multi-server federated architecture, namely, a graph federated learning architecture (GFL). However, existing work assumes a fixed amount of data at clients and the training of a single global model. This paper proposes a decentralized online multitask learning algorithm based on GFL (O-GFML). Clients update their local models using continuous streaming data while clients and multiple servers can train different but related models simul-taneously. Furthermore, to enhance the communication efficiency of O-GFML, we develop a partial-sharing-based O-GFML (PSO-GFML). The PSO-GFML allows participating clients to exchange only a portion of model parameters with their respective servers during a global iteration, while non-participating clients update their local models if they have access to new data. In the context of kernel regression, we show the mean convergence of the PSO-GFML. Experimental results show that PSO-GFML can achieve competitive performance with a considerably lower communication overhead than O-GFML.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130823811","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"On Performance Limit for Bandwidth-Constrained Distributed Parameter Estimation Systems","authors":"Alireza Sani, A. Vosoughi","doi":"10.1109/CISS53076.2022.9751162","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751162","url":null,"abstract":"We consider a bandwidth-constrained distributed parameter estimation problem, where each sensor makes a noisy observation of an unknown random source $theta$. Each sensor is unaware of $theta$'s prior distribution and the actual dynamic range of its observation, and simply assumes that its observation is limited to a finite interval [$tau_{k}, tau_{k}$]. Each sensor quantizes its observation using a multi-bit uniform quantizer, where the quantization step size is chosen according to $tau_{k}$. Sensors send their quantized observations to a fusion center (FC), that is tasked with estimating $theta$ based on the received data from the sensors. We derive the Bayesian Fisher information, which is the inverse of the Bayesian Cramer-Rao lower bound, for two types of random $theta$, namely Gaussian and Laplacian $theta$. To quantify the amount of information loss on $theta$ when the FC uses the quantized observation for estimating $theta$, due to both limited dynamic ranges at the sensors and uniform quantization, we examine the derived Fisher information at the asymptotic regimes, when the quantization rate and $tau_{k}$ go to infinity. We also provide two accurate approximations of the Fisher information for two cases of (i) binary and (ii) high rate multi-bit quantizers. Through simulations we explore the conditions under which the information loss on $theta$ is negligible and demonstrate the accuracy of the provided approximations.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126479517","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Industrial Application of Multi Target Detection in Thermal Images from Different Cameras with DeepLearning","authors":"Berkan Unutmaz, Isiotan Erer","doi":"10.1109/CISS53076.2022.9751159","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751159","url":null,"abstract":"In this study, the main aim is to automatically perform the manual target detection process used in the camera field of view testing of mass-produced thermal cameras. A data set is prepared by taking images using different mass production cameras and different test systems. With this prepared data set multi target detection architecture is proposed. This proposed hybrid architecture consist of ResNet50 block, which is used for feature extraction, and YOLOv3 block. The accuracy of this proposed architecture to detect targets whose number and position changes in each image, compared with Minimum Output Sum of Squared Error(MOSSE), Single Shut Detection(SSD), Aggregate Channel Features(ACF), Recurrent Convolutional Neural Network(RCNN), FAST-RCNN, FASTER-RCNN, and YOLO versions target detection architectures. As a result of this comparison, it is seen that the proposed hybrid architecture has higher accuracy than other architectures. The use of proposed hybrid architecture in the camera field of view test of each camera produced with mass production will reduce the workload and increase the accuracy of the camera field of view calculation.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122230439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimal Drug Dosing to Achieve the Desired Actual Neutrophil Counts (ANC) in Chemotherapy Induced Myelosuppression","authors":"V. Radisavljevic-Gajic","doi":"10.1109/CISS53076.2022.9751192","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751192","url":null,"abstract":"In this paper we have first considered the well-known myelosuppression mathematical model of Lena Friberg and her coworkers from the system analysis point of view and study the linearized model steady state stability, controllability, observability, and scaling of model variables. We have found that the original model has poor stability properties at steady state since all its eigenvalues are very close to the imaginary axis. Using theory of multi-time scale systems, it was noticed that the linearized dynamics of two state variables is slow (corresponding to the numbers of maturing cells in the third compartment and the number of circulating cells) and that three remaining state variables display fast dynamics (corresponding to the number of proliferative cells and the number of maturing cells in the first and second compartments). In order to avoid numerical computations with large numbers scaling of system state variables by a factor of 109 has been utilized. In the second part of the paper, a method was proposed for optimal chemotherapy dosing using a result from optimal control theory in order to reduce the amount of administrated chemotherapy drugs and to keep the number of neutrophil cells above a pre-specified desired ANC (actual neutrophil count) level. It was shown that in the case of continuous dosing, the variable optimal amounts of the drug have to be administrated daily based on feedback information regarding the actual count of neutrophils. This result mathematically establishes that administrating constant amount of drugs daily cannot provide the optimal dosing schedule. The obtained results open a door for modern personalized and optimized medicine that requires daily monitoring of fundamental variables and daily drug administration in variable quantities based on the actual state of the patient's fundamental variables (parameters) for the considered decease.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126031313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Safety Control for Prime Focus Spectrograph","authors":"Ting-Han Fan, Athindran Ramesh Kumar, P. Ramadge","doi":"10.1109/CISS53076.2022.9751172","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751172","url":null,"abstract":"Prime Focus Spectrograph is a telescope system to be deployed in Hawaii. The system consists of roughly 2400 controllable observation units, which we call the cobras, and the cobra control problem is to reach the targets while avoiding collisions. We decompose this problem into cobra assignment and trajectory planning. The cobra assignment adopts an efficient near-optimal algorithm that maximizes the target acquisition and completely avoids final-position collisions by losing only 0.9% of the target acquisition. The trajectory planning uses a collision-based search which lowers the in-transit collisions to almost zero.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"516 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133167159","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"On Explainability of A Simple Classifier for AR(1) Source","authors":"Cem Benar, A. Akansu","doi":"10.1109/CISS53076.2022.9751181","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751181","url":null,"abstract":"The heuristic reasoning and experiments based design approach have been the pillars of studies on artificial neural networks. The explainable network performance is required for most applications. We focus on a simple classifier network for the two-class case of AR(1) data sources. We trace the input statistics through the network and quantify changes to explain relationship between accuracy performance, optimized parameters and activation function types employed for the given architecture. We present test accuracy results for various network configurations with different dimension and activation types. AR(1) source model for a two-class case is utilized to generate training and test data sets of the experiments due to its ease of use for analytical study. We quantify the relationships with well known metrics among signal (class) statistics, network architecture, activation function type and accuracy for several correlation coefficient pairs of the two AR(1) sources utilized in this paper. It is observed from the experiments that the analyses of data, input-output relationships of hidden and output layer nodes for the given architecture provide invaluable insights and guidance to judiciously design a neural network and to explain its performance based on characteristics of the building blocks.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127728459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Massive-MIMO Channel Capacity Modeling for mURLLC Over 6G UAV Mobile Wireless Networks","authors":"Xi Zhang, Qixuan Zhu, H. Poor","doi":"10.1109/CISS53076.2022.9751152","DOIUrl":"https://doi.org/10.1109/CISS53076.2022.9751152","url":null,"abstract":"Due to significant improvements in throughput and efficiency, the massive multiple input multiple output (MIMO) has been recognized as a key component of the massive ultra-reliable and low-latency communications (mURLLC) over the emerging sixth generation (6G) wireless networks to support the continuously increasing mobile user's traffic demands. Leveraging the dynamic architecture of unmanned aerial vehicle (UAV) swarm, it is executable to deploy a massive MIMO communication networks between the UAV swarm and a ground-station (GS) equipped with massive antennas. In this paper we integrate the UAV technique with massive MIMO wireless communications and derive the channel capacity performance for the UAV-to-GS massive MIMO channel. We develop the massive MIMO channel model from a group of UAVs, equipped with single antenna, to the GS, equipped with uniform rectangular antenna array, and obtain the closed-form expressions for channel capacity. Finally, we use numerical results to validate and evaluate the channel performances of our proposed massive MIMO communications.","PeriodicalId":305918,"journal":{"name":"2022 56th Annual Conference on Information Sciences and Systems (CISS)","volume":"18 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128636175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}